Workshops & Courses
研讨会与课程
W-33
Workshop
Meta-analysis and reproducibility: publishing fast and following scientific research ethics
Saturday, June 27th, 2026
Organizer(s):
Zakari Sissou, University of Parakou; Gbadamassi Dossa, Xishuangbanna Tropical Botanical Garden; Denis Njoroge, Technical University of Dresden
Description
Nowadays, most universities and research institutes still follow the "Publish or Perish" policy as a means of evaluating their scientists and students. As such, publications inevitably become a core criterion of assessment for students, lecturers, and researchers for obtaining degrees or getting promoted. Most of the time, publications and h-index are used as evaluation criteria of the scholarships or grants. In the case of academic staff, these policies are regulated by the evaluation committees of their institutions and government agencies. Scope: Primary data in biophysical and social research generally come from very costly and time-consuming research efforts. Most primary research requires special funding and laboratory work to obtain data. These research funds are highly competitive, and many students and early-career researchers end up not obtaining research funding, which would consequently delay their graduation as they would not be able to meet at least the publication requirements. In addition, primary data generally produce several other datasets from different regions in a similar field, which could be directly targeted by data collectors and are therefore usually underutilized even when available. Compared to primary local research results, meta-analyses and review articles provide synthesis of multiple primary results and allow for expanding future directions in a given research field. Goals: The goal of this workshop is to equip researchers (most early career and students) with the practical skills and conceptual understanding necessary to conduct rigorous, reproducible meta-analyses that enhance the reliability of scientific synthesis. Objectives: at the end of this workshop, the participants are expected to: Understand the difference between a systematic review and a meta-analysis, and to formulate a structured research question using the PICO/S framework to guide a reproducible evidence synthesis. Know and practice the calculation of common effect sizes (e.g., Hedge's g, odds ratios) and perform a basic random-effects meta-analysis using statistical software to generate forest plots and quantify heterogeneity. Develop a pre-registered protocol for a meta-analysis and apply key reproducibility practices, such as creating a reproducible analytical pipeline and assessing publication bias, to enhance the transparency and reliability of findings Know how to get deposit data and analysis code in repository with DOI (digital object identifier) number assigned to these. Know how to have a well reproducible workflow that will enable them to meet nowadays criteria from journals for data and analysis.
Program Outline
SESSION 1: INTRODUCTION TO METADATA ANALYSIS
Target audience: Lecturers and researchers
Duration: 2 hours
Objectives: Participants are able to:
Understanding the different types of reviews and the importance of conducting a review
Differentiation of the conditions of qualitative and quantitative review
Understanding the theory behind the “ Effect -size”
Expected results: Participants will be able to:
Differentiate between types of review and understand and explain the importance of conducting a review
Choosing between a qualitative and quantitative review
Manually calculate an “ Effect -size”
Contents / description:
Module 1
1.1 Importance of metadata analyses (meta-analyses and systematic reviews)
Module 2
1.2 Types of reviews and characteristics
Module 3
1.3 Good practices for meta-analysis and systematic review
Module 4
1.4 Quantitative or qualitative review (meta-analysis or systematic review)
Module 5
1.5 What is an “ Effect -size”?
1.5.1 Case study 1: Measures of “Effect size “ for dichotomous variables
1.5.2 Case study 2: Measures of “Effect size “ for quantitative variables
Session’s methods and tools:
Theoretical Lecture
Group work
Training manual
SESSION 2: STEPS IN CREATING A REVIEW
Target audience: Lecturers and researchers
Duration: 2 hours
Objectives: The participant must be able to:
Formulate a research question and a research topic for a meta-analysis
Understanding inclusion and exclusion criteria and searching for articles in databases
Filter study information from the downloaded data, then extract data from the filtered and downloaded studies.
Expected results: The participant is able to:
Formulate a research question and a research topic for a meta-analysis
Understanding inclusion and exclusion criteria and searching for articles in databases
Filter study information from the downloaded data, then extract data from the filtered and downloaded studies.
Contents / description:
Module 1
2.1 Formulation of questions
2.1.1 Formulating a research question
2.1.2 PICO Formula
2.1.3 Some problems related to the wording of the questions
Module 2
2.2 Establishment of the protocol
2.2.1 Principles
2.2.2 Key Components
Module 3
2.3 Article Search
2.3.1 Article search strategies
2.3.2 Literature search
Module 4
2.4 Selection of items
2.4.1 Good practices
2.4.2 Use the “CRAAP” test
2.4.3 Existing tools for evaluating published articles
Module 5
2.5 Data Extraction
2.5.1 Best practices for data extraction
2.5.2 Data Extraction Tools
Session’s methods and tools:
Theoretical Lecture
Group work
Practical Session 1: Discussion on the formulation of questions and the choice of topics
Practical Session 2: Formulating keywords for article search
Practical Session 3: Introduction to R
Practical Session 4: Literature search in databases and thematic statistics
Practical Session 5: Filtering Titles and Summaries
Practical Session 6: Data Extraction
Tools needed:
Excel
R software
RStudio
Install “metafor”, “ggplot2” and “tidyverse”
Internet browser and access to Scopus and Web of Science.
SESSION 3: DATA ANALYSIS FOR META-ANALYSIS
Target audience: Lecturers and researchers
Duration: 3 hours
Objectives: The participant must be able to:
Understanding the meaning and calculation of “ Effect -size”
Fixed-effect and Random-effect models
Understanding the heterogeneity of the “ effect size” of studies and the role of moderators in meta-analysis
Present the results as a graph in the forest and perform a sensitivity analysis
Expected results: The participant will be able to:
Define and calculate the “ Effect -size”
Fixed-effect and Random-effect models
Define the heterogeneity of the “ effect size” of the studies and suggest moderators for a meta-analysis
Disseminate the results of the data analysis in the form of a graph in the forest and calculate a sensitivity analysis
Contents / description:
Module 1
3.1 Calculating the “ Effect -size” in R
3.2 Missing data and data imputation
Module 2
Fixed-effect and Random-effect models?
3.2.1 Source of variation
3.2.2 Heterogeneity
Module 3
3.3 Forest plot and Funnel plot graphs
Module 4
3.4 Sensitivity Analysis
Session’s methods and tools:
Theoretical Lecture
Group work
Practical Session 1: Designing your database for R meta-analysis
Practical Session 2: Calculating the “Effect -size” using R scripts
TP 3: Heterogeneity analysis and calculation of the “summary” moderators’ “effect -size” using R scripts
Practical Session 4: Forest plots and meta-regressions
Tools needed:
Excel
“metafor” package
“ggplot2” and “tidyverse”
SESSION 4: REPRODUCIBILITY IN SCIENCE
Target audience:
Duration: 2 Hours
Objectives: To enable participants to:
.Get familiar with reproducible science
Get to know important tools available for reproducible sciences (R markdown, Quarto, version control)
Get familiar with Github and Git.
Get code and data uploaded on Zenodo
Expected results:
Able to create a project with a good reproducible workflow
Able to link computer/R studio/Posit with Github account
Contents / description:
Module 1
4.1
Session’s methods and tools:
Theoretical Lecture
Group work
Practical Session 6: navigating through Git, R Markdown, Github,
Practical Session 7: Get code and data uploaded into Zenodo with obtention of doi number
Etc.
Materials that participants need to bring:
Apart from having a laptop and notebook, participants must install the followings:
Software Engage digitizer (https://digitizer.sourceforge.net/)
RStudio (last version)
R version 4.5.1
R Packages metafor, litsearch r, ggplot 2, tidyverse
